Table 1.
EEG studies of motor recovery.
| References | N, Stroke patients | Stroke acuity at EEG collection | EEG data collection, timing design | N, EEG channels | Patient State during EEG recordings | EEG connectivity technique | Networks/regions investigated | Clinical scores collected | Main findings |
|---|---|---|---|---|---|---|---|---|---|
| Giaquinto et al. (23) | 34 | Subacute | Longitudinal | 16 | Resting | PSA | Bilateral hemispheres | BI | In patients with the greatest recovery, inter-hemispheric EEG balance increased over time. |
| Gerloff et al. (30) | 11 | Chronic | Cross-Sectional | 28 | Behavioral-Triggered | PSA | Bilateral hemispheres, sensorimotor cortex focus | MRC, manual muscle testing | In patients who recovered well from capsular stroke, connections within motor were reduced in the stroke/lesioned hemisphere but increased in the contralesional hemisphere. |
| Kaiser et al. (31) | 29 | Subacute to chronic | Cross-sectional | 61 | Behavioral-triggered | ERD, ERS LC | M1s | ESS, MRC, MAS | Motor impairment correlated with contralesional ERD. Lesioned ERD correlated with spasticity. Lesioned ERS correlated with both motor impairment and spasticity. |
| Fallani et al. (27) | 20 | Subacute | Cross-sectional | 61 | Behavioral-triggered | SWI, imaginary coherence analysis during rest and MI | Bilateral hemispheres | FMA | Lesioned hemispheres showed a reduction in SWI scores and reduced local efficiency. Inter-hemispheric imbalance related to greater motor impairment. |
| Wu et al. (32) | 12 | Subacute to chronic | Longitudinal | 256 | Resting | PSA | Interhemispheric connections between M1s, lesioned connectivity in sensorimotor network | FMA-UE. | At baseline, connectivity in lesioned M1 is marker of motor status. Increase in connectivity in lesioned M1 biomarker of motor recovery. Lesioned M1–SMA connectivity increased and M1–parietal connectivity decreased in parallel with motor gains, |
| Bönstrup et al. (33) | 12 | Acute to chronic | Longitudinal | 64 | Behavioral-Triggered | PSA | Sensorimotor network | FMA-UE NHPT, grip strength | Initial up-regulation of brain activity after stroke correlates with neuronal reorganization for post-stroke recovery |
| Pichiorri et al. (28) | 28 | Subacute | Cross-sectional | 61 | Behavioral-triggered | PSA, PDC | Bilateral hemispheres | FMA-UE | Post-BCI MI training desynchronized alpha and beta activity, which correlated with motor improvement. |
| Thibaut et al. (34) | 55 | Chronic | Cross-Sectional | 128 | Resting | PSA | Bilateral frontal, central and parietal networks | FMA | Patients with balanced interhemispheric beta activity experienced greater motor function recovery. |
| Philips et al. (35) | 30 | Chronic | Longitudinal | 58 | Behavioral-Triggered | GMA and Network Based Analysis | Bilateral hemispheres | FMA-UE | Reduced contralesional intradensity and high initial values of local lesioned efficiency predicted better motor recovery. |
| Agius Anastasi et al. (36) | 10 | Subacute | Longitudinal | 32 | Behavioral-triggered | BSI | Bilateral hemispheres | Motricity index, FMA | Baseline BSI higher in stroke and more pronounced in the cortical stroke and predicted FMA. |
| Chen et al. (37) | 37 | Subacute | Cross-sectional | 32 | Behavioral-triggered | DCM, PSA | SMA and bilateral M1s | WMFT, FMA-UE, | Beta plus gamma or theta network features predicted good recovery. |
| Pichiorri et al. (38) | 30 | Subacute | Cross-sectional | 64 | Resting | PDC-connectivity | Sensorimotor network | TMS-CST integrity, European stroke scale and FMA | Inter-hemispheric coupling was higher in patients with preserved CST integrity. Lower sensorimotor beta coupling correlated with clinical impairment. |
| Vecchio et al. (39) | 139 | Acute | Cross-sectional | 27 | Resting | SWI | Bilateral hemispheres | NIHSS, BI, and ARAT | NIHSS, Barthel, and ARAT scores correlated with SWI. Baseline gamma SWI predicted final NIHSS. |
| Eldeeb et al. (40) | 3 | Chronic | Longitudinal | 15 | Behavioral-triggered | PDC-based network connectivity | Sensorimotor cortex | FMA-UE, grip strength. | An NIBS intervention led to improvement in PDC; improvements in PDC correlated with improvements in hand function. |
| Bönstrup et al. (41) | 30 | Chronic | Cross-sectional | 64 | Behavioral-triggered | PSA | Lesioned parietofrontal motor network | UEFM, NHPT, grip strength | Parietofrontal coupling was stronger in stroke patients and correlated with residual motor impairment. |
| Saes et al. (42) | 21 | Chronic | Cross-sectional | 64 | Resting | PSA | Bilateral hemispheres | FMA-UE | Stroke patients showed higher BSI scores between M1s, with activity differences most pronounced in delta and theta frequency bands. In the delta and theta bands, BSI negatively associated with FM-UE. |
| Bönstrup et al. (43) | 33 | Acute to subacute | Longitudinal | 64 | Behavioral-triggered | PSA | SMA, M1 | FMA-UE, NHPT, grip strength | Acute stroke–lesioned brains failed to generate the LFO signal. LFOs progressively increased at 1 and 3 months. Re-emergence of the LFO correlated with motor recovery. |
| Bartur et al. (44) | 14 | Subacute | Cross-sectional | 64 | Behavioral-triggered | ERD | Bilateral M1s | FMA, BBT | Lesioned ERD positively correlated with residual motor function and the magnitude of EMG in the hand. |
| Cassidy et al. (45) | 62 | Acute, subacute, chronic | Longitudinal | 256 | Resting | PSA | Interhemispheric connections between M1s and intra-hemispheric motor connections | FMA-UE. | Greater coherence between inter-hemispheric delta M1 activity correlated with poorer motor status. Decreases in inter-hemispheric coherence between lesioned M1 and contralesional M1 correlated with better motor recovery. |
| Romagosa et al. (46) | 36 | Acute to subacute | Cross-sectional | 16 | Resting | BSI, LC | Bilateral hemispheres | FMA, BBT, NHPT, MOCA, BI. | BSI correlated with FMA-UE, but not with FMA-LE. Laterality coefficient correlated with FMA-UE and FMA-LE. |
| Kawano et al. (47) | 40 | Subacute | Cross-sectional | 19 | Resting | Phase synchrony index | Inter-hemispheric connections between M1s and intra-hemispheric motor connections | FMA-UE | The inter-hemispheric motor cortical alpha-band PSI was lower in stroke patients and correlated with UEFM. Contralesional central theta-band PSI was higher in patients, and correlated with improvements in FMA-UE. |
| Hoshino et al. (48) | 24 | Subacute | Cross-sectional | 5 | Behavioral-triggered | Amplitude envelop correlations | Frontocentral motor areas | FMA | Bilaterally higher intrahemispheric and interhemispheric activity at 4 weeks predicted higher and lower limb function at 8 weeks. |
| Saes et al. (49) | 39 | Acute | Cross-sectional | 62 | Resting | PSA | Bilateral hemispheres | FMA-UE | Baseline BSI theta values predicted greater upper limb motor impairment 6 months after stroke. |
ARAT, Action Research Arm Test; BI, Barthel Index; BSI, Brain Symmetry Index; Chronic, 6 months and beyond; CST, Corticospinal Tract; DCM, dynamic causal modeling; ERD, Event-related desynchronization; EMG, Electromyography; ERS, Event-related synchronization; FC, functional connectivity; FMA, Fugl Meyer Assessment; GMA, generalized measure of association. Acute; 0–2 weeks after stroke; LC, laterality coefficient; LE, Lower Extremity; BBT, Box and Block Test; LFO, Low-Frequency Brain Oscillations; M1, Primary Motor Cortex; MI, Motor imagery; NIHSS, National Institute of Health Stroke Scale; PDC, Partial Direct Coherence; PSA, Power Spectral Analysis; pts, patients; Subacute, 2 weeks-6 months after stroke; SWI, Small World index; UE, Upper Extremity.